© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 41-GEMMASCOPE-2-RES-262K
    7. 64807
    Prev
    Next
    INDEX
    Explanations

    multi-head attention

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post_all/layer_41_width_262k_l0_big
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    xadmin
    0.42
     పొ
    0.41
     Posting
    0.40
    seqNum
    0.40
     जुड़े
    0.39
    atakan
    0.39
     Sequence
    0.38
    繋
    0.38
     enumeration
    0.38
     急
    0.37
    POSITIVE LOGITS
     subspace
    0.80
     parallel
    0.70
     subspaces
    0.69
     independently
    0.68
     parallèle
    0.59
    parallel
    0.58
     paralelo
    0.56
     independent
    0.55
    Parallel
    0.55
     Parallel
    0.52
    Activations Density 0.002%

    No Known Activations